[ntp:questions] Reasons of NTP not to use GPS source

David Taylor david-taylor at blueyonder.co.uk.invalid
Tue Sep 17 09:51:20 UTC 2013


On 17/09/2013 09:05, Igor Pavlov wrote:
> Thanks, Brian.
>
> I'll do all these things.
> Now will try to fix problem with PPS: it's level 3.3V, and serial port
> seems to not recognising such low level.
> Now playing with time2 parameter and my GPS now stoped getting "x".
>
> I can't understant one thing.
> Should system time be fluently corrected by NTP so in some term offset of
> active ("*") source would be close to 0?

You can check whether the serial port is seeing the 3.3V PPS signal with 
my program:

   http://www.satsignal.eu/software/net.htm#SerialPortLEDs

Note that a PPS signal of at least several microseconds duration is 
required by the PC, and perhaps 100 milliseconds or more to be visible 
on my program.  Most GPS use 100-200 ms for the pulse, which is adequate 
for seeing seen by eye.  NTP would need to be stopped while running my 
program.  On checking the manual for your device:

   http://www.geostar-navigation.com/file/geos1/User_Manual_GeoS-1M_en.pdf

it seems that the maximum length of the PPS pulse is only 2 
milliseconds, and the default is just 1 millisecond, so that may not 
show on my program, but should be seen correctly by the PC using the DCD 
pin.

Yes, the offset for the selected source should be close to zero (+ or 
-).  Checking on my Windows stratum-1 PCs right now:

XP: offset 0.086
Win-7/64: offset -0.032
Win-8/64: offset -0.009

There is nothing "special" about those values, just an indication of the 
sort of offsets you might see.  My FreeBSD and Linux stratum-1 PCs show:

FreeBSD: -0.001
Raspberry Pi 1: 0.002
Raspberry Pi 2: 0.001
Raspberry Pi 3: 0.001
Raspberry Pi 4: -0.001 (user-mode PPS)

The offsets on Windows PCs being synced from the Internet will be /much/ 
larger, perhaps more than 10 milliseconds.
-- 
Cheers,
David
Web: http://www.satsignal.eu



More information about the questions mailing list